Re: [Xen-devel] Xen 4.11.1 panic

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 19.12.18 at 13:54, <JBeulich@xxxxxxxx> wrote:
>>>> On 19.12.18 at 12:55, <bouyer@xxxxxxxxxxxxxxx> wrote:
>> On Wed, Dec 19, 2018 at 04:05:57AM -0700, Jan Beulich wrote:
>>> In any event, both Andrew and I must have overlooked the one
>>> crucial place due to which the assertion is indeed wrong from
>>> put_page_from_l2e():
>>> 
>>>         int rc = _put_page_type(pg, false, mfn_to_page(_mfn(pfn)));
>>> 
>>> Not allowing for preemption there is fine if the L2E is pointing to
>>> an L1 table, but is now wrong if the L2E points to another L2,
>>> which surely is the case when you see the assertion trigger.
>> 
>> Should we just change false to true here, or should the cases above be 
>> handled differently ?
> 
> Switching from false to true here is just the initial part of the
> necessary change - if you did just this, you'd end up hitting
> the ASSERT() right after the line above. There's quite a bit
> more to it, and it needs to be done pretty carefully.

Actually there was no reason to alter the free_l2_table() paths
in the XSA-273 fixes: A switch to shadow mode can only occur
when validating page tables. Therefore I think you could safely
revert the respective hunks, which includes deleting the
ASSERT() you found triggering.

Jan



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel